A model to estimate losses due to bovine mastitis for Argentinian dairy herds

• Published in: Journal of animal science Vol. 94; p. 589
• Main Authors: Richardet, M, Solari, H, Vissio, C, Bartolome, J, Bo, G, Turiello, P, Bogni, C, Larriestra, A
• Format: Journal Article
• Language: English
• Published: Champaign Oxford University Press 01.10.2016
• Subjects: Animal diseases; Cattle; Computer simulation; Culling; Demography; Disease control; Economic models; Estimates; Evaluation; Infections; Literature reviews; Mastitis; Microorganisms; Milk; Mimicry; Reproduction; Reproductive failure; Stochasticity; Tissues; Transition probabilities; Upgrading
• ISSN: 0021-8812; 1525-3163
• DOI: 10.2527/jam2016-1224

A comprehensive economic evaluation of disease control implies developing models to capture the complexity and dynamics of the production system, especially for diseases like mastitis, which has multiples effects such as milk losses, increased risk of culling or a higher likelihood of reproductive failure. Objective: to describe preliminary results of estimated clinical (CM) and subclinical mastitis (SCM) frequency caused by S. aureus and their milk associated losses by a stochastic simulation model. Methodology: The model simulates discrete events overtime mimicking a real Holstein herd in terms of production and reproduction. The system has been divided into compartments involving reproduction, production, disease, feeding and culling/mortality events and their respective costs overtime. The model has been written in C language and its parameters have been gathered through a literature review. The model starts with a user defined herd in terms of demography and health status. From that point, the system projects the whole dynamics of the herd for a specific time horizon (e.g., 12 mo). The model focuses on S. aureus infections and drives the infection within the herd considering transition probabilities among different cows (uninfected or subclinically or clinically infected). The system updates the whole herd and disease information every 2 wk. As an example, the model has been run 100 times in a 200-cow herd. Results: The annual projection showed a median gross CM prevalence of 3% (q1 = 2%; q3 = 4%) and a median gross SCM prevalence of 21% (q1 = 17%; q3 = 25%). Estimated milk losses due to CM and SCM were 2.87 and 1.40 l/cow/day, respectively. All results are consistent with observational data recently published in Argentina. The model evaluation and verification on relevant assumptions need to be done. The model runs satisfactorily and it can be customized for the user. Further development will involve the inclusion of multiples contagious and environmental microorganisms.